Lab Log

The Legendary Team: a Claude Engineering Squad That Works Claude Code ships with one agent by default. That’s fine for isolated tasks. It breaks down the moment you have a system with multiple layers — API, frontend, database, CI — because no single context window can hold the whole thing without losing precision at the […]

Build an Autonomous AI Coding Agent on One PC Running a multi-agent coding swarm on a single machine sounds like something that needs a server rack. It doesn’t. On one Ubuntu box with an RTX 3090 and 64 GB RAM I had three coordinated Claude Code instances writing, reviewing, and testing code concurrently — no […]

Welcome to WordPress. This is your first post. Edit or delete it, then start writing!